Displaying 4 results from an estimated 4 matches for "qgpd".
Did you mean:
gpd
2007 Jun 06
0
A question about riskmeasures() vs. qgpd() in library(evir)
Dear List,
This inquiry probably does not directly pertain to R.
I am using library(evir) to learn EVT. Based on my reading of things,
it is my understanding that if one wants to calculate quantiles of
GPD, one could use either riskmeasures() or qgpd(). However, using
data(danish) as an example, the quantile estimates produced by
riskmeasures() are considerably different from those produced by
qgpd() as follows:
> library(evir)
> data(danish)
> out <- gpd(danish, 10)
> riskmeasures(out, c(0.999, 0.9999))
p quant...
2007 Jun 19
2
Function -return value
...the two parameters in parab.
parameter <- function(v) {
v1 <- v[v>mean(v)+0.5*sd(v)]
v2 <- v[v<mean(v)-0.5*sd(v)]
alpha=min(v1)
beta=max(v2)
para <- fitgpd(v1,alpha, method="pwmu")$param
parab <- fitgpd((-v2), (-beta), method="pwmu")$param
v1.fit <- qgpd(ppoints(v1, a=0.5), alpha, para[1], para[2])
v2.fit <- qgpd(ppoints((-v2), a=0.5), (-beta), para[1], para[2])
alpha
beta
para
parab
}
tapply(variable, list(a, b),parameter)
I would be grateful if anyone can give me some advice. Many thanks
--
View this message in context: http://www.nabble....
2008 Sep 18
0
Joint distributions
...om this distribution.
I know the distributions f(Hs) and f(Tp), which I have assumed to be
generalized Pareto distributed as
library(ismev)
library(evd)
Zh=5; Zt=10;
Tppars=gpd.fit(x,Zt)$mle
Hspars=gpd.fit(y,Zh)$mle
and so the return values of each parameter separately are calculated as:
yTp100=qgpd(1-p1/100,Zt,Tppars[1],Tppars[2])
yHp100=qgpd(1-p1/100,Zh,Hppars[1],Hppars[2])
My question are as followed:
1) Is there any function in R which can help me calculate a joint
distribution when the separate distributions are known?
or
2) Is there any way I can extend the contour lines in my plot to c...
2007 Jun 13
2
Fitted Value Pareto Distribution
I would like to fit a Pareto Distribution and I am using the following codes.
I thought the fitted (fit1) should be the fitted value for the data, is it
correct? As the result of the "fitted" turns out to be a single value for
all.
fit=vglm(ycf1 ~ 1, pareto1(location=alpha), trace=TRUE, crit="c")
fitted(fit)
The result is
fitted(fit)
[,1]
[1,] 0.07752694